Morgan Stanley Deploys OpenAI Chatbot to Assist 16,000 Financial Advisors
Morgan Stanley is rolling out an OpenAI-powered chatbot to its financial advisors, giving them instant access to a vetted internal research database. The tool aims to make every advisor as knowledgeable as the firm's top strategists while implementing strict guardrails to prevent AI errors.
Morgan Stanley is rolling out an advanced chatbot powered by OpenAI's GPT-4 technology to help its 16,000 financial advisors access the bank's massive repository of research and data. The firm currently tests the artificial intelligence tool with 300 advisors and plans a wider rollout in the coming months. This initiative makes Morgan Stanley one of the first major financial institutions to implement a practical, internal generative AI tool following the viral success of ChatGPT.
The chatbot acts as an on-demand research assistant, allowing advisors to instantly find answers drawn from approximately 100,000 pieces of vetted Morgan Stanley content. A firm executive compares the experience to having the bank's chief strategy officer sitting next to the advisor during a client call. By limiting the AI's knowledge base to pre-approved internal documents rather than the entire internet, the bank aims to provide highly accurate and relevant information.
To address the known risks of generative AI, such as convincing but incorrect responses or "hallucinations," Morgan Stanley implements strict user guardrails. The bank relies on human testers who actively attempt to break the platform to identify potential flaws before the wide release. With these ongoing monitoring processes and the use of high-quality, curated information, the wealth management division expresses confidence in the tool's reliability.